> So, in summary, it's a glibc bug that has been closed as "WORKSFORME" and
> will never be fixed [3].
>
> In the test-pthread-rwlock test, we cannot just use
> PTHREAD_RWLOCK_WRITER_NONRECURSIVE_INITIALIZER_NP, because the *purpose* of
> the test is to check the behaviour of the rwlocks with the POSIX-specified
> API, not with some alternative API.
What we could do, probably, is:
1) Define a module 'pthread-rwlock-writer-prefer', that guarantees to
solve the first two readers-writers-problems [1] (see also [5]).
2) Optionally, implement fair rwlocks, using one of the algorithms
described in [2][3][4], in a module named 'pthread-rwlock-fair'.
3) If module 'pthread-rwlock-writer-prefer' is not in use and
m4/pthread_rwlock_rdlock.m4 has determined that readers are preferred
and the number of CPUs is > 4, skip the test.
Bruno
[1] https://en.wikipedia.org/wiki/Readers%E2%80%93writers_problem
[2] https://arxiv.org/abs/cs/0303005
[3] https://arxiv.org/abs/1309.4507
[4] https://rfc1149.net/blog/2011/01/07/the-third-readers-writers-problem/
[5] https://stackoverflow.com/questions/2190090/
